The conventional wisdom that Thomas's votes follow
Antonin Scalia's is reflected by Linda Greenhouse's observation that Thomas voted with Scalia 91 percent of the time during October Term 2006, and with Justice
John Paul Stevens the least, 36% of the time.
[97] Statistics compiled annually by Tom Goldstein of
SCOTUSblog demonstrate that Greenhouse's count is methodology-specific, counting non-unanimous cases where Scalia and Thomas voted for the same litigant, regardless of whether they got there by the same reasoning.
[98] Goldstein's statistics show that the two agreed in full only 74% of the time, and that the frequency of agreement between Scalia and Thomas is not as outstanding as is often implied by pieces aimed at lay audiences. For example, in that same term, Souter and Ginsburg voted together 81% of the time by the method of counting that yields a 74% agreement between Thomas and Scalia. By the metric that produces the 91% Scalia/Thomas figure, Ginsburg and Breyer agreed 90% of the time. Roberts and Alito agreed 94% of the time.
[99]